FACILITATING AGRICULTURE AUTOMATION USING STANDARDS Robert K. Benneweis P. Eng Outline

Available standards Developing standards Implemented automation Standard based automation

implementation Potential standard based

automation implementation Potential tractor – implement

control Automation safety Facilitating Standard

implementation Future agriculture automation Conclusions

Available standards ISO 11783

ISO 11783 specifies a serial data network for control and communications on forestry or agricultural implements

It standardises the method and format of data transfer between sensors, actuators, control elements, display and storage units

ISO 11783 provides interoperability and interchangeability of electronic units between different types of implements and implements from different manufacturers

Specifies a standard electronic interface that is similar to other standard tractor – implement interfaces, such as: Three point hitch

standard – ISO 730, ISO 789, ISO 2332

Hydraulic remote connection – ISO 5676, ISO 17567

PTO standard – ISO 500

ISO 11783 consists of the following 14

parts: Part 1 – General (FDIS) Part 2 – Physical layer (IS) Part 3 – Data Link layer (IS) Part 4 – Network layer (IS) Part 5 – Network management

(IS) Part 6 – Virtual terminal (IS)

(DAmd)

Part 7 – Implement message application layer (IS) (DAmd)

Part 8 – Power train messages (IS)

Part 9 – Tractor ECU (IS) Part 10 – Task controller and

management information system data interchange (FDIS)

Part 11 – Mobile data element dictionary (FDIS)

Part 12 – Diagnostics (DIS) Part 13 – File server (FDIS)

Part 14 – Automated functions (WD)

Network Structure (ISO 11783)

Parts 1 to 5 and 12 specifies a standard ISO 11898 based data communication network for connected agriculture mobile systems

Parts 1 to 5 and 12 are the ‘foundation’ of the standard on which controls are able to be implemented as specified by the remaining parts

Part 6 specifies a standard operator interface for setting up and operating an implement or sensor system

Part 6 is for ‘manual’ operations of implements or for possible set up of automated functions

Parts 7, 8, 9, 10, 11 & 14 specify requirements to implement automation Part 7 specifies measure

and command messages for control of: PTO Hitch Auxiliary hydraulic

valves Guidance Task controlling Tractor operation

control Part 8 specifies measure

and command messages for control of: Engine

Transmission Part 9 specifies a network

interconnect unit between the implement bus and the tractor bus. It also specifies 3 classes of tractors: Class 1: provides

messages – Basic sensor data, power management & light controls

Class 2: provides messages

– Date & time, hitch draft, aux valve status, distance traveled & direction

Class 3: provides messages

– Hitch, PTO & Aux valve commands

Location (NMEA 2000) messages

Front hitch option Power control

Part 10 specifies task controlling and the interface between the management computer and the implement:

• Data interchange schema based on XML

• Definitions of tasks to be performed including site specific control

• Defines tractor – implement configuration and offsets between connection points and implement elements

Part 11 is the data dictionary for part 10 data entities:

Part 14 specifies measure and command messages for control of:

• Headlands implement control of tractor functions

• Other automation function of implements

Parts 13 specifies

requirements for a file server on the mobile implement system

Other Available standards IEC 61162-3

Maritime navigation and radio communication equipment and systems - Digital interfaces – Part 3:

International version of the USA National Marine Electronic Association’s NMEA 2000 standard

ISO 15077 Operator controls — Actuating

forces, displacement, location and method of operation

Requirements for operator controls associated with virtual terminals, as specified in ISO 11783-6, are given in Annex B

ISO 15765 Diagnostics on CAN Primarily intended for

diagnostic systems, developed to also meet requirements from other CAN based systems needing a network layer protocol

SAE J1939 Recommended Practice for a

Serial Control and Communications Vehicle Network for light and heavy duty on and off road vehicles and equipment

Developing standards ISO 25119 ISO 25119 is being developed by

TC23/ SC19/ & SC3 JWG6 Working draft documents

based on IEC 61508 are being prepared

ISO 25119 provides safety requirements and guidance on the principles for the design of high risk functional parts of control systems used in agricultural and forestry machinery to ensure human safety

It applies to high risk functional parts of electrical/ electronic/ programmable electronic systems and as part of mechatronic systems

It does not specify which dangerous critical functions and which categories shall be used in a particular case

ISO 25119 consists of 4 parts

Part 1: General principles for design and development

Part 2: Concept phase Part 3: Series Development,

Software, Hardware Part 4: Production,

Operation, Modification

Part 1 specifies the phases of the overall equipment lifecycle which are: Phase 1: Concept

phase Phase 2: Series

Development, Software, Hardware

Phase 3: Production, Operation, Modification, Decommissioning

ISO 25199-Part 1 specifies the phases of the overall equipment lifecycle

ISO 25119-Part 1 specifies risk analysis to determine hardware and software design requirements

Wireless sensor networks

A new standard is being developed by TC23/ SC19/ WG5 for wireless sensor networks

The standard specifies wireless data communication between remote sensors, base stations and/ or interfaces to wired networks for application in agriculture, including communication to management systems

The wireless remote sensors can be mounted on tractors, implements, stationary equipment and facilities, in fields and green house crops, on animals, with products and in the environment

Wireless sensor are able to replace expensive wire connections, or for locations subjected to high wear and/ or not accessible by wire connection (i.e. rotating shafts)

Implemented available automation Headlands control

An individual controller records operator control actuations as the tractor – implement completes a turn between two passes in a field such as: Throttle down/ gear

down as approaching headlands

Stop machine primary function

Reduce machine secondary function such as fan rpm

Raise machine from in ground to headland positions

Complete turn to next pass

Lower machine from headland position to in ground

Start or resume machine functions

Throttle up/ gear up

Stand alone headlands controller has individual connections to operator controls and machine actuators

Map based rate control

Site specific control of implement operations Planned implement

operations and settings based on geo-referenced or site specific locations within a field using an management computer

The planned commands are transferred to a map based controller on the tractor – implement connected system

Navigation (gps) receiver is connected to the map based controller to provide the actual location of the tractor – implement in the field and the map based controller sends the planned commands for that actual location to the implement controllers

Actual implement operations and settings are sent to map based controller to be logged for that actual location to a file for transfer to office computer

Temporal specific control of implement operation Planned implement

operations and settings based on seasonal timing

Available map based controllers use proprietary connections & signals

Auto steering Straight line paths

Navigation (gps) receiver is connected to the auto steering system to provide the actual location of the tractor – implement in the field

Implement operating width is entered into the auto steering system

Tractor – implement location of an operator controlled first straight line pass within a field is logged in the steering controller system

The steering direction of each subsequent pass is controlled by the auto steering system using the navigation (gps) receiver location information and the previous logged pass offset by the implement operating width

Operator controls headlands turn from one pass to next pass

Curved line paths Operations same as

straight line path, except first pass can be curve

Auto steering system has capability to determine desired direction of travel from previous curved path offset by the implement width

Available auto steering system use proprietary connections & signals

Standard based automation Headlands control

Headlands control can be implemented using ISO 11783 messages and controllers

Part 7 tractor facilities messages can be used for implement – tractor control

Part 14 can be used to record events and synchronise controls Throttle down/ gear

down as approaching headlands

Stop machine primary function

Reduce machine secondary function such as fan rpm

Raise machine from in ground to headland positions

Complete turn to next pass

Lower machine from headland position to in ground

Start or resume machine functions

Throttle up/ gear up No other proprietary H/W

controllers are required

Headlands control

Map based rate control

Site specific control can be implemented using ISO 11783-10 messages and a task controller

Use Part 11 Data Dictionary data entity definitions to identify message data

The task controller sends and receives process data messages to/ from ISO 11783 compliant implement controllers

No other proprietary H/W controllers are required

Map based rate control

Auto steering

Straight line or curved path steering control can be implemented using ISO 11783-7 messages and a compliant steering controller

Steering controller connected to the tractor bus

Navigation controller connected to ISO 11783 implement bus

ISO 11783-7 steering commands the navigation controller are sent to tractor ECU for transmitting to the steering controller

Auto steering

Combined auto steering & headlands control Auto steering possible using

ISO 11783 Headland control possible using

ISO 11783 Both auto steering and

headlands control can be

implemented together on ISO 11783 network

It is therefore possible to have auto steering when turning from the first pass to the next pass

Therefore the operator does not have to turn the tractor – implement through the headlands turn

Combined auto steering & headlands control

Potential standard based automation implementation Implement controlled tractor

ISO 11783 Part 7 messages for implement control of tractor functions PTO measured status

and activation commands

Hitch measured status and activation commands

Auxiliary hydraulic valves measured status and activation commands

Guidance or direction status and commands

ISO 11783 Part 7 tractor command requests and control Tractor cruise control Combine constant PTO

speed and cruise control with guidance

Auxiliary valve slip control and cruise control with guidance

ISO 11783 Part 14 automated functions Messages between

tractor & implement controller & automation

master controller on the implement

Logging sequence based on time, distance or event triggers

Automation control by messages from function controllers triggered by logged sequence of events

Advantage of implement controlled tractors Reduced time to

accomplish same area operations as manual operations

Reduce fuel consumption for same area operation as manual operations

More accurate implement operations, i.e. depth control, application rate control

Reduce implement cost by direct use of auxiliary hydraulic valves on tractors

Examples of standard based implement control of tractors Tractor auxiliary valve

control to maintain

desired implement tillage depth

Tractor auxiliary valve control to maintain implement fan rpm

Tractor speed control to maintain desired draft and depth

Other possible implement control of tractor supplied functions

Planned tractor – implement control

ISO 11783 is a communication network, control is by messages

With map based rate control, pre-planned application rates are communicated to an implement controller

The implement controller then completes the rate commands when specified

The tractor – implement connected system can be controlled by the same method

Work in the field can be pre-planned for the most efficient operations

Tractor – implement path planning is one example of this pre-planning of work

Planned tractor – implement control

Robots

With auto steering, headland control, implement control of tractor, map based rate control and path pre-planning, is the operator really required?

Can a tractor- implemented connected system operating autonomously?

Yes, demonstrated as early as 1958 at Reading University using an International Harvester B250 tractor

An autonomous self-propelled windrower was successfully operated in the late 1990

Demeter

Sponsored by NASA, New Holland and Carnegie Mellon University demonstrated an autonomous self propelled windrower, which operated through night without an operator

Why didn’t Demeter make it to market? It was a research project

and required cost effective development to meet market requirements

The market was not ready for an autonomous implement

According to one manager from New Holland it lacked a sensor

• A lawyer sensor! Clones

It will be difficult for autonomous farm equipment to be accepted by farmers

Farmers will still want to have the operator controlling the tractor

With automation it is possible to have one operator control a number of implements to achieve greater operational efficiencies

Using wireless networks between standard implement networks, a master implement with an operator can ‘lead’ a number of similar autonomous implements

During harvest, it is also possible to have an autonomous grain cart follow a combine during hopper unloading, then travel to unload in a large truck at the edge of the field

Herds/ Packs

After operating with clones, it will possible to move to a master autonomous implement (robot) leading a number of similar robots

This group of autonomous implements can operate together without a leader, cooperating together doing their assigned work, similar to a herd of animals or a pack of wolves

These autonomous implements can be small (i.e. a single row planter or a single row harvester) but in large numbers to accomplish the work of a large implement

With is arrangement, if one small implement malfunctions, it does not stop the work of the other implements

Loading or replenishing the bins or hoppers on planters or fertilisers can be completed one implement at a time, therefore not stopping the work of all the implements

Autonomous implement safety Operator safety

W.S. Reid presented a paper at the 2004 ASAE conference in Ottawa Canada titled ‘Safety in Perspective, For Autonomous Off Road Equipment’

Data was presented that indicated the 68% of farm accidents in regions of NA involved tractors (45%), machines (12%) or trucks (11%)

For all tractor accidents, rollover was 43%, run over was 18% and fall off was 10% equaling a total of 71% of all accidents

A significant number of accidents involve the operator being caught in an attached operating implement (left cab with implement operating)

Autonomous implements removes the operator or passenger from the tractor which would reduce these accidents by the above %

Robotics safety Obstacle avoidance systems

Reid proposed a zone of safety around the connected implement

Operations would only proceed if the zone of safety was not occupied by an individual or obstacle

This zone of safety can be observed by vision, radar or laser systems

A similar type of vision system was implemented on the Demeter windrower

Facilitating Standard implementation AUTOSAR (Automotive Open

System Architecture) The primary objective of

AUTOSAR is to create an infrastructure which encourages cooperation on E/E standards while maintaining competition on innovative applications

AUTOSAR vision is an improved complexity management of highly integrated E/E architectures through an increased reuse and exchangeability of SW modules between OEMs and suppliers

The AUTOSAR project goals will be met by specifying and standardizing the central architectural elements across functional domains, allowing industry competition to focus on implementation

The AUTOSAR solution is based on standardized SW interfaces which support both the exchangeability of SW components and HW independence

Agrosar (Agriculture Open System

Architecture) Propose a partnership similar to

AUTOSAR to enable system-wide process optimization of agriculture electronic systems (i.e. partitioning and resource usage) and allow for local optimization to meet the runtime requirements of specific devices and hardware constraints

AGROSAR can improved complexity management of highly integrated E/E architectures through an increased reuse and exchangeability of SW modules between OEMs and suppliers

The AGROSAR solution is based on standardized SW interfaces which support both the exchangeability of SW components and HW independence

A possible AGROSAR software architecture

Future agriculture automation

Ray Kurzweil in his book ‘The Age of Spiritual Machines’ provides possible future capabilities of agriculture automation

2019

Computers are embedded everywhere, walls, desks, clothing, bodies

$4000 (1999 $) computer have the computational capability of the human brain

Massively parallel neural nets and genetic algorithms are used

Automated driving systems are highly reliable and have been installed in nearly all roads

Using the above predictions agriculture could have fully automated

field operations have widespread

availability of bioengineering technology, but the danger of creating diseases

have computer programs determining crop rotations and nutrients to be applied for forecasted weather conditions

2029

$1000 (1999 $) computers have the computational capacity of about 1000 human brains

Computing now conducted on massively parallel neural nets based on reverse engineering of the human brain

Nanoengineered robots have the microbrains with the capacity of the human brains

Using the above predictions then agriculture could be fully automated have no human

employment for production

2099

The reverse engineering of the human brain is complete

Substantial advantages to machine based intelligence

There are basic rights of machine based intelligence

??? Conclusions Standard based systems assist with

the development of agriculture automation

Provides the ‘foundation’ of a control and communication serial data network

Messages and protocol now available for implementing automation

Standard based systems allow the

combining of implemented single functions

Auto steering and headlands combined to provide non operator control of tractor – implement system operations

Standard base systems facilitate the

rapid implementation of autonomous agriculture equipment

Effort can be focused on the autonomous operations

Efforts are not need on a base system to support autonomous operations